A service to expose HuggingFace's T5-Base model as an API
- docker 20.10+
- Minimum RAM size needed to run is 8GB, Recommended is 16GB+
- To use less resource as well and for convenience sake the application relies uses the CPU-support version of the transformers library
transformers[torch]
. As the result, a GPU is not necessary for building and running the application.
# Clone repository to your local and navigate into the folder
git clone https://github.com/horlali/api.translator-t5-base
cd api.translator-t5-base/
# Build and start the application with docker-compose
docker-compose up --build
See links to model features, documentation, and training proceduces here